The Shanghai Artificial Intelligence Laboratory, in cooperation with Ruijin Hospital affiliated to Shanghai Jiao Tong University School of Medicine and other institutions, launched the medical multi-modal basic model group "PU Medical 2.0" at the "2023 Healthy China Sinan Summit". Compared with the previous version, "PU Medical 2.0" has significantly improved model scale, data modality and functions, added models in multiple fields, expanded language parameters, and covered medical images and medical texts. Multiple data modalities such as and biological information have greatly expanded its application scope. What's more worth mentioning is that this release also includes a model evaluation module, which provides a reference standard for the evaluation of medical AI model capabilities, making it easier for developers to better evaluate and optimize model performance.
The Shanghai Artificial Intelligence Laboratory, Ruijin Hospital Affiliated to Shanghai Jiao Tong University School of Medicine and other partners released the medical multi-modal basic model group "PU Medical 2.0" at the "2023 Healthy China Sinan Summit". The new version adds multi-domain models and incremental language parameters, covering multiple data modalities such as medical images, medical texts, and biological information. It also adds an evaluation module to provide a reference for medical model capabilities. This release realizes one-stop open source of large medical model groups, providing capability support for cross-domain, cross-disease, and cross-modal AI medical applications.
The one-stop open source of "PU Medical 2.0" brings great convenience to the research and development and application of medical AI. It marks that China has made important progress in the field of medical AI and provides a solid foundation for the innovative development of future medical technology. the basis of. Its cross-modal, cross-field, and cross-disease application capabilities will greatly promote the application of medical AI technology and benefit more patients.